A hybrid optimized error correction system for time series forecasting

• Published in: Applied soft computing 2020-02, Vol.87, p.105970, Article 105970
• Main Authors: de Oliveira, João Fausto Lorenzato, Pacífico, Luciano Demetrio Santos, de Mattos Neto, Paulo Salgado Gomes, Barreiros, Emanoel Francisco Spósito, Rodrigues, Cleyton Mário de Oliveira, Filho, Adauto Trigueiro de Almeida
• Format: Article
• Language: English
• Subjects: Artificial neural networks; Extreme learning machines; Hybrid intelligent systems; Particle swarm optimization; Time series forecasting
• ISSN: 1568-4946; 1872-9681
• DOI: 10.1016/j.asoc.2019.105970

Time series forecasting is a challenging task in machine learning. Real world time series are often composed by linear and nonlinear structures which need to be mapped by some forecasting method. Linear methods such as autoregressive integrated moving average (ARIMA) and nonlinear methods such as artificial neural networks (ANNs) could be employed to handle such problems, however model misspecification hinders the forecasting process producing inaccurate models. Hybrid models based on error forecasting and combination can reduce the misspecification of single models and improve the accuracy of the system. This work proposes a hybrid system that is composed of three parts: a) linear modeling of the time series, b) nonlinear modeling of the error series, and c) combination of the forecasts using three distinct approaches. The system performs a search for the best parameters of the linear and nonlinear components, and of the combination approaches. Particle swarm optimization is used to find suitable architecture and weights. Experiments show that the proposed technique achieved promising results in time series forecasting. •Extreme learning machines with linear activation functions are used as linear models.•The optimization searches the best parameters for linear and nonlinear combination.•The hybrid system can perform three types of combination: sum, linear and nonlinear.•Linear combination (sum with weight) is initially proposed in the present work.